Will Apple’s Anti-Algorithm Save Music?

Beats-1-Zane-Lowe-Ebro-Darden-Julie-Adenuga

A look into the magic of the first global college radio station

Beats 1 Radio is the most contrarian bet in the music world right now. And that’s exactly why it’s exciting.

The progress in the music + tech space in recent years (if you can call it progress) has been dominated by an obsession with algorithms. The promise has been that algorithms, if properly developed and fed with data, would be able to offer infinite personalization that would make bigger music fans of us all.

But it hasn’t really worked out that way. Despite the best efforts of the bots, having an infinite supply of music at our fingertips has made our music listening more provincial… not less. Our biological hard wiring to favor the familiar leaves us with little immediate affection for anonymous recommendations, and the vast majority of music listeners have little appetite to put in the effort associated with finding new music. To this challenge, most technologists reply by saying that the algorithms just need more: more data inputs, more technological honing, more time.

Apple seems to believe that we’ve been approaching this thing wrong all along. And so, with Beats 1, the biggest technology company on the planet has introduced the ultimate anti-algorithm. And I think this radically contrarian bet is one of the biggest reasons to be excited about music’s future.

I find Beats 1 so exciting for three reasons.

It’s College Radio Manned by Megawatt Celebrities

One of the most surprising aspects of Beats 1 is how incredibly and endearingly unpolished it is. The unscripted spontaneity of nearly every show gives the station the handcrafted vibe of a late night on college radio. There are genuine jokes and honest technical mistakes, songs stopped mid-stream and interviews where truly nobody knows what’s going to be said next.

But here’s the rub: the DJs on this ramshackle college radio broadcast are a collection of the biggest stars in the music world. Beats 1 features everyone from Eminem to Elton John to Dr. Dre and Drake… and that’s just in the first week. But when these superstars leave their PR people behind and open up through music, the result is the most honest into these individuals that I’ve ever heard. It’s addictive, irresistibly shareable stuff, and has led some to look at Beats 1 as the future of Twitter.

It’s Simultaneously Global and Local

Much has been made of the fact that Beats 1 broadcasts live to 100 countries simultaneously (well, at least when it’s not on replay). But, while conceptually interesting, the fact that Beats 1 is broadcasting to 99 other countries doesn’t in and of itself have much value to your average listener.

But what makes the global nature of Beats 1 interesting is the degree to which the DJs are repping their cities: Ebro does a show that is of NYC, not just from it. And when Julie Adenuga says “this is London,” she actually delivers on it. While this local aspect is just a glimmer right now, it holds promise for a truly exciting vision of what it could mean to be a global radio station. Benedict Evans has said that Apple Music reminds him of Google Maps as it provides “manual curation at scale.” With Beats 1, we can take the analogy of Google Maps one step further, as it holds the promise of allowing anyone, anywhere to peek into another part of the globe and appreciate it in a new way. Powerful stuff.

It’s Actively Passive

The fatal flaw made by so many in the music streaming space is that they massively over-estimate the amount of effort your average listener is interested in expending.   They create services for their 25 year-old music nerd selves, and lose track of the fact that the continued dominance of FM radio and the enormously broad appeal of Pandora is rooted in the incomparable ease of these two services (oh, and they’re free).

Beats 1 not only eliminates complication, it prevents it. The only things you can do with it are turn it on and turn it off. But the hyperactive curation built into the broadcasts makes Beats 1 feel like a very active experience though it requires absolutely nothing from the user. It’s as effortless as FM radio, yet often as exciting as a mixtape discovery.

These three reasons makes Beats 1 an incredibly massive experiment in what radio and digital music can be.  About a year and half ago, I wrote that “Radio has to radically re-think what it needs to be.”  At that time, it was more of a lament than an actual statement of hope.  But now, with Beats 1, Apple has given us reason to think that good old fashioned radio might help pave the way into the future of music.

Did Pharrell Just Kill the Exclusive?

pharrell-freedom-680x676I’m pretty sure I’ve found the song of the summer, and it just so happens that this same song has killed off a central strategic pillar of several music streaming services.

First of all, let’s start with some background.  Pharrell’s “Freedom” is an irresistibly infectious hit (la la la la… la la la la la—good luck getting that out of your head).  And, blurry legal issue aside, Pharrell’s been on an uninterrupted A-list hit streak for a few years now.

As well, Apple spent a lot of time and money to let people know that he was going to drop an exclusive on Apple Music to launch the whole enterprise putting it everywhere from press releases to their big TV spot  to the social channels of everyone from Apple to Pharrell himself.

Now we’re a week and a half into the track, and it’s really starting to catch the mainstream with everyone from the USA Today to Kanye talking about it.

But here’s the rub: now that the track is catching on everywhere, it’s available everywhere.  As soon as everyone started to like the exclusive, the exclusivity was gone.

This raises some pretty fundamental questions not just about this effort from Apple but also about music exclusives overall.

First, in a world where even hit songs from huge artists take at least a week to really catch hold, is there any benefit to a brand or streaming service to have it for the first few days when a track is just starting to take off?  Does your average (Spotify) user care?  More pointedly, is there anyone who does care about having the track in those first few days who won’t go through the trouble of finding it on YouTube or another not-so-legal destination?

Second, what is the residual value of an exclusive track?  In a month’s time, will anyone remember that Apple had the track exclusively for a few days?  Or is the only value realized if you keep doing exclusives over and over again, leaving you chasing the proverbial rabbit around the race track to make any of it matter?

Third, how much of your brand do you need to uncomfortably contort to promote exclusives?  Apple found this out the hard way– Zane Lowe played the song repeatedly during the time that the song was exclusive to Beats One radio… but the trouble is, that’s the opposite of the point of Beats One Radio, and completely contrary to pretty much everything Zane Lowe stands for (Twitter complained plenty about this).  To promote the exclusive, Apple had to compromise the core beliefs of the service.  Tricky.

These three questions clearly demonstrate that Apple’s last week with Pharrell casts some pretty big doubts on the exclusive windows that are presumed to be a pillar of the strategies of streaming services.

Can Taylor-esque boycotts leave marks on Spotify and scare Eddy Cue?  Yes, but these scenarios are limited to the nearly singular kinds of artists that yield as much market power as she does (see below for the source of Taylor’s singularly stunning market power).

CI3savEUkAAFpwU

To be fair to Apple, exclusives aren’t a strategy of theirs as much as they are an enabler of their curation strategy.  Apple has bet big on curation over algorithms, but their experience with Beats Music taught them that anonymous curation doesn’t work: playlists made by anonymous editors won’t work among mainstream audiences… but playlists made by celebrities just might.  So, for Apple, celebrity exclusives are an ingredient that turbo charges their curation strategy.

But, for others (read: Tidal), the implications are far more worrying.  From the first moments of their re-launch, Tidal has ignored the high fidelity sound that had been their niche trademark, and bet big on exclusives of everything from songs to videos to playlists.  In fact, the only place you can currently stream Prince tracks is Tidal.  But you didn’t know that until just now, and that’s the point.

One of the biggest stars on the planet just invalidated one of the few strategic paths that have been identified by streaming services. If not exclusives, what should they do?

More on that in the days to come.

Get Rich and Die Trying: A Look Inside the Music Industry’s Collective Death Wish

Money in fist

When you love music as much as I do, it’s hard not to be troubled when you see the entire industry careening toward economic collapse. It’s even more troubling when it’s so evident that the industry is bringing its demise upon itself.

The clubby collaboration between the major record labels and Spotify has placed the entire music industry on top of a precarious investment bubble, and in so doing has left the industry on the teetering brink of economic implosion.

Spotify’s most recent round of funding a few months ago valued the company at $8.4 billion. The entire US recording industry was valued at $6.972 billion at the close of 2014. Though Spotify has yet to make a profit (their net loss tripled from 2013 to 2014), its valuation is based upon the growth of its customer base. It’s valued like a tech company, not a traditional music or entertainment company. One might think this is healthy for the industry, a sign that the music industry has embraced technology and is receiving valuations that acknowledge as much.

But here’s the first problem.

One of the reasons why tech valuations can be supported is economies of scale: as a customer base grows, a company not only gains a greater hold on the market but its costs go down (as calculated per customer).  But that is not at all the case with Spotify.  Sure, there are some tiny tech-based economies of scale as it relates to Spotify’s servers and whatnot.  But nearly 75% of their post-tax revenue is paid out to record labels, and these royalty costs do not go down at all as Spotify grows.  As Spotify grows, more songs are listened to, and their cost base goes up.

Here’s the second problem.

Spotify’s insistence on the freemium model locks it into chasing customers that just aren’t profitable.  A paying Spotify customer generates 26 times more revenue than a user of the free service, and although the percentage of Spotify users is stuck at around 25%, paying customers account for 91% of the company’s income.

But won’t the freemium model convert users to paying customers over time?

Here’s the third problem.

The average customer spends about $48 per year on iTunes.  Given iTunes scale, you can consider their average customer to be a pretty good proxy for an overall average customer (probably even spendier than your average customer, given that they are customers who are still buying music to begin with).  So, if you are betting on people to convert to the paid tier of Spotify and other streaming services, you are expecting your average person to nearly triple their yearly expenditure on music.  And you are expecting people to do this despite the fact that there are going to be legal and free alternatives available to them… with the second most popular alternative (second to YouTube) provided by Spotify itself!

For paid (read: profitable) streaming services to succeed, the pricing is probably going to need to come down by half. And, as it turns out, Apple tried exactly this. In the early days of their negotiations with labels, Apple had planned to charge $5 per month. But the major labels wouldn’t allow this. Subsequently, Apple tried to dig in its heels at a $7.99 monthly charge, but once again the major labels wouldn’t sign any agreements below $9.99 per month.

Why are labels not trying to steer this boat away from what is obviously an iceberg of permanent unprofitability?

And here’s the fourth (and biggest) problem.

The major record labels, collectively, own about 20% of Spotify.  That means that they have a stake in Spotify that is worth about $1.7 billion, and they have seen this stake double in value just since September of 2013.  So, despite the fact that Spotify is unprofitable and is contributing mightily to the persistent unprofitability of the entire music business, why on earth would the labels care?  They are getting significant royalty payments that Spotify has to pay regardless of profitability, and with their collective 20% stake in Spotify they are riding the valuation that is predicated upon the false hope that the company will some day become profitable.

This intertwined interdependency of the music industrial complex risks ruining the economics of the music industry permanently.

So how can the industry innovate and pull out of this death spiral? This very question is the fuel for future blog posts.

Stay tuned.

Hello, Again.

In years past, the passage of 18 months would mean that you’d be listening to different music. Now, eighteen months fundamentally changes how you’re listening to music.

Over the course of the past year and a half, I’ve spent many of my work days up close with many of the topics on which I used to opine on this blog. Paradoxically, I’ve been too busy to write at a time when I’ve had plenty to say.

With this realization, I’ve resolved to get back to writing.

For those of you who used to visit, some things will be the same. I’m still the same guy who has been obsessed with music since before he could remember. My enthusiasm for foisting musical recommendations on anyone and everyone hasn’t waned.

But, as they should, some things will change. I won’t be posting (m)any mp3 links, because my music behavior has changed along with most everyone’s.  And my writing will probably lean more toward the professional than the personal, as much of what I’m going to cover will relate to topics that I’m tackling in my professional career (albeit without all the confidential bits).

So let’s see if I can get back to merging the professional and the personal in ways that you’ll find interesting.

As always, thanks for reading.

John

Not Writing and Writing

Screen Shot 2014-05-13 at 4.11.08 PM

It has been a very quiet few months here at the blog.  Part of this has been due to the natural ebbs and flows of work/life busy-ness that has led to prior lapses of authorship.  But another part of this absence is that I have been writing; I’ve just been doing it elsewhere.

Many of you know that I have been doing more and more writing and speaking over the past year or so, giving talks to various companies and classrooms who have been kind enough to invite me.  The discipline instilled by these speeches (i.e., I’m getting on stage and need to have something to say) has spurred me to distill and refine a lot of my thoughts on how the future of marketing innovation can be inspired by the re-invention of the music industry.

Since many of the ideas in these speeches started as posts on this blog, I thought it was a good time to re-unite these two worlds.  Here are the slides for a speech that I’m giving this Friday at the Kellogg School of Management.  Though you won’t be able to view the videos and whatnot, you’ll still get a sense of things.

I hope you enjoy it.  As I’ll still be fine tuning things for the next few days, shoot me a note with any thoughts you might have.

And more to come soon on the pages of this blog.  Stay tuned.

 

Bringing My A Game for the Algorithm

securedownloadEveryone wants to look good in front of a mirror, even if it’s a virtual one.

My experience with Beats Music yesterday triggered an unexpected pang of performance anxiety.  I began, as I always do, by looking through the list of recommendations served up “just for me” by the Beats Music algorithm.  But as I made my way through this tour, I found myself surrounded by mile markers of all too obvious influences.  

Front and center was far more hip hop than I would self-describe.  Thankfully, the usual suspects were comfortably couched by a fair amount of old school credibility.  While I couldn’t begin to guess when I last listened to Eric B & Rakim (or will again), my inner music snob found it nice to have them around.

On either side of this rap heap were towering totems of (recent) life phases.

On one side there was a lineup that I’ve carried with me from my collegiate years, along with each and every other person who went off to school in the mid-90’s.  As I scrolled past Pearl Jam and Smashing Pumpkins and Jane’s Addiction, I began to quicken the pace as I searched for the recommendations that would validate the peerless music knowledge that I (thought I) had during those years.  Where was the Mary Lou Lord b side?  How about the rare Pearl Jam bootleg that would prove that even though I liked the biggest band on the planet at the time I did so in a clever way?  Alas, the only faint nod that Beats gave to my collegiate self-estimation was an Air album.

Flanking these juggernauts of college rock was the soundtrack from nights of my twenties.  Browsing lounge dwellers such as Portishead and Basement Jaxx and DJ Shadow, I could almost feel myself staying up past midnight once again.

This walk down my musical memory lane was impressive in its accuracy and unnerving in its honesty.

Nowhere to be found was the rare jazz cut, the emerging artist, or the instrumentalist from a country’s whose name I can’t quite spell.  While I’d like to explain this away by blaming the inherent logic of every recommendation engine (built to please, not surprise), I knew I couldn’t let myself off this easy.  Because behavior doesn’t lie.  Algorithms serve up what you have demonstrated that you will enjoy: not necessarily what you’d like to tell others that you enjoy.

The unexpected wrinkle within the algorithms that are working their way into more and more facets of our everyday lives is the fact that we’re going to care about how we look like in all of these virtual mirrors.  We may even find our decisions swayed by the desire to shape the recommendations that will get played back to us.  Will your next Netflix viewing be a decision made for you, or a decision made for what will get reflected back to you?

Here’s a track that I’d like to see the next time I glance in the Beats Music mirror.   I think you’ll enjoy it as well.

Two, Antlers

Remembify

UnknownI found the early days of Netflix to be incredibly frustrating.  Despite my earnest attempts to allow their servers to get to know me, I kept finding myself being unwillingly shoved to the far edge of the long tail.  Even when I tried to convince them that I did indeed appreciate my fair share of mindless entertainment, I got volley after volley of French documentaries and forgotten TV series foisted upon me.  Although there was (is?) a part of me that wouldn’t mind being thought of as someone who appreciates French documentaries, forgotten in the entire process was the fact that I was there to be entertained.

I later learned that Netflix’s behavior was largely financially motivated (as they paid less for the obscure), but they are far from alone in romanticizing the distant end of the long tail.  The most recent service that promises to enrich your life by serving you up the incredibly obscure is Forgotify.   Inspired by recent statistics released from Spotify that evidence that 20% of the songs in their catalogue (a full four million) have never been played once, Forgotify will serve you up a song that has never been heard.

While it’s a cute premise, Forgotify’s promise is tantamount by helping you decide what to cook for dinner by suggesting a list of foods that nobody has ever eaten.  Even putting aside the fact that the reason why these songs/foods haven’t been consumed is the simple fact that they’re not worthy of consumption, Forgotify is trying to solve the wrong problem.

The real issue isn’t that forgettable music is being forgotten by the world; it’s that we don’t remember much of what we have already discovered to be awesome.  If Forgotify wanted to help me, it would tap into that track that made my week two years ago, but quickly found its way to the recesses of my hard drive.

I’ve written about this topic before, but it’s worth a quick re-hit.  Take the track below, “Play Your Part (Pt. 2) by Girl Talk (remember him?).  I’ve played this song over 20 times since downloading it (a non-trivial amount), but the last time I played this song was nearly a year ago.  Had I not made a point to look at what I was listening to a year ago, who knows if and when I’d ever enjoy this again.

Play Your Part (Pt. 2), Girl Talk

I always want to be exposed to the best of the new music in the world, and I’ll always want help with that.  But, equally, I want help remembering what I already love.  Both are equally important to any music lover.